Results 1 to 6 of 6
My .xinitrc file:
Code:
exec fluxbox &
fbpid=$!
sleep 3
{
gdeskcal &
kopete &
adept_notifier &
dcop kded kded loadModule knemod &
} &
wait $fbpid
The bold line ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
- 04-23-2008 #1Linux Newbie
- Join Date
- Apr 2007
- Posts
- 214
How to start knemo from .xinitrc on fluxbox?
My .xinitrc file:
Code:exec fluxbox & fbpid=$! sleep 3 { gdeskcal & kopete & adept_notifier & dcop kded kded loadModule knemod & } & wait $fbpid
The bold line was what was supposed to start knemo from the startup, but for some reason it does not, differently from the other ones.
The same command, however, executes it, from the command line or from fbrun.
- 04-23-2008 #2Linux Guru
- Join Date
- Nov 2007
- Location
- Córdoba (Spain)
- Posts
- 1,513
It's just a guess, but dcop depends on a running instance of dcopserver, and maybe on such a early stage it can't initialize for some reason. You can try to add a delay with sleep 10; dcop <whatever>
You can also try to launch dcopserver implicitly from this file. I don't really know. In any case, can you see any error from knemod if you look into the logs?
- 04-24-2008 #3Linux Newbie
- Join Date
- Apr 2007
- Posts
- 214
The sleep 10 didn't work.
As I'm in an awkward stage of intermediate newbie, I don't knew (and still don't know) where to get any relevant log.
I got knemo working by adding kdeinit to the script, however:
Code:exec fluxbox & fbpid=$! kdeinit sleep 3 { gdeskcal & kopete & adept_notifier & dcop kded kded loadModule knemod & } & wait $fbpid
- 04-24-2008 #4Linux Guru
- Join Date
- Nov 2007
- Location
- Córdoba (Spain)
- Posts
- 1,513
Nice that you got it working, either way

Since you are using .xinitrc as your config file, I assume you use startx to launch X, don't you?
If that assumption is true, then you can look at the output by just going back to the vt from where you ran startx.
You can just redirect the output to a file. This is the most convenient way, because this way, you can then just open a term on your x session, and use tail to get the logs directly in front of your face.
You could just use "startx > ~/mylogfile.txt" to log the output to that file, but I usually just do it in my xinitrc this way:
My xinitrc:
Then, in an xterm, you could do "tail -f ~/logs/fvwm.log" to see the log as it's updated.fvwm &>~/logs/fvwm.log 2>&1
- 04-24-2008 #5Linux Newbie
- Join Date
- Apr 2007
- Posts
- 214
Thanks for the tip. I think I'll make my line somewhat like yours. Just exiting will often add many more message lines and as far as I know, there's no way to scroll back to read what got past the screen.
But I don't understand what is the whole "2>&1" part. (but I just searched for that between the quotation marks in google and there are some answers around)
I thought that just what was before that part would be enough... (even though I thought that something like this would somehow redirect even the graphic output to the text file and end in a blank screen, but apparently I'm wrong
)
...
Fluxbox has its own "-log" option, but it logs different things, not the whole "background" stuff.
- 04-24-2008 #6Linux Guru
- Join Date
- Nov 2007
- Location
- Córdoba (Spain)
- Posts
- 1,513
You don't have to exit, you can do control+atl+f1 to change to vt1, f2 for vt2 and so on. Usually, X is started on vt7, so to go back to X, just use control+alt+f7. You can use then shift+page up/down keys to scroll a bit.
Search for "bash redirection" insteadBut I don't understand what is the whole "2>&1" part. (but I just searched for that between the quotation marks in google and there are some answers around)
That line I posted, in fact, has a slight inoffensive error (well, more like it's redundant). It should be
In unix/linux, the default streams, stdin, stdout and stderr are referenced using three numbers, respectively: 0, 1 and 2.Code:fvwm > ~/logs/fvwm.log 2>&1
So, the > redirects stdout to a file, the 2>&1 part redirects stderr to stdout (which in turn is redirected to a file). This way, both stderr and stdout will be logged to a file.
This is the way that sh does it, so, it's *the way* to do it if you are going to use this config on shells other than bash.
Bash also accepts this form:
This redirects both (stderr and stdout) to a file. That's why I said above that the first line I posted has a small typo on it which made it redundant.Code:fvwm &> ~/logs/fvwm.log
Yep, it was a typo, as I explain above.I thought that just what was before that part would be enough...
The graphic stuff is not in shell land. It's done via a framebuffer or via an accelerated graphics driver which operates in a completely different way, so it can't be saved this way.(even though I thought that something like this would somehow redirect even the graphic output to the text file and end in a blank screen, but apparently I'm wrong
)


Reply With Quote

